RSI hype builds: where AI self-improvement stalls, per TMTPost analysis
A TMTPost analysis says recursive self-improvement (RSI) has quickly moved from lab shorthand to industry buzzword, with OpenAI, Anthropic and Google DeepMind repeatedly citing progress, and with Anthropic and OpenAI warning about loss-of-control risks and urging slower development. In China, Zhipu founder and chief scientist Tang Jie disclosed the company's first RSI-oriented engineering practice and likened the core technique of its next "fully self-training" model GLM-6.0 to RSI; Xiaomi called its MiMo-V2.6 series "a key step in exploring RSI"; and Alibaba CEO Wu Yongming said the Qwen team is exploring RSI with some progress. Practitioners quoted in the piece argue that because humans still set goals and make final decisions, these efforts look more like an extension of existing technical paths than the start of something new.
RSI is shifting from a lab term into part of vendor and investor narratives, yet there is still no shared definition or verifiable standard for whether it marks a real capability jump or just deeper automation of existing workflows.
(本文作者为 飞向TAI空,钛媒体经授权发布)

图片来源:unsplash
RSI(递归自我改进)火了。
这个聚焦于AI自我演进的概念,似乎在业界、学界连定义和标准都来不及统一的情况下,就快速进入了落地阶段。
近半年来,OpenAI、Anthropic、Google DeepMind等前沿AI实验室和初创公司频频提及RSI方面的进展。尤其9月中旬以来,Anthropic和OpenAI数次预警RSI的失控风险,甚至呼吁放缓研发进度,为安全保障争取时间。国内方面,RSI的技术演进和判断也更多走向台前,智谱创始人兼首席科学家唐杰9月17日披露公司首个RSI方向的工程化实践,并将下一代“完全自训练”模型GLM-6.0的核心技术定位类比为RSI;小米则在9月22日发布MiMo-V2.6系列模型,称之为“探索RSI的关键一步”;阿里巴巴集团CEO吴泳铭则在云栖大会演讲中提到,目前人们摸索出通往超级人工智能的具体技术路径就是RSI,他还透露,Qwen团队正在进行RSI方面的探索,并取得了一定的进展。
#OpenAI#Anthropic#Google DeepMind#Zhipu AI#GLM-6.0#Xiaomi#MiMo-V2.6#Alibaba#Qwen#Meta#Claude#Weco AI